A resource-capability perspective

We draw on and extend the resource-capability-based view (hereafter RBV for short) to form the theoretical base of the paper. According to the RBV, a firm consists of bundles of resources and capabilities it needs to produce and sell a good or service (Penrose, Reference Penrose1959; Barney, Reference Barney1991, Reference Barney2001). Resources are defined as ‘the tangible and intangible assets’ (Barney, Reference Barney2001: 54), and ‘they are stocks, not flows’ (Teece, Reference Teece2010a: 689). In contrast, capabilities are defined as the capacities ‘to utilize resources to perform a task or an activity against the opposition of circumstance’ (Teece, Reference Teece2014b: 14). As such, capabilities ‘flow from the astute bundling or orchestration of resources’ (Teece, Reference Teece2014a: 14). Capabilities are intrinsically intangible and are undergirded by resource orchestration processes and practices. Recent development in the RBV emphasized the importance of capabilities to sustained competitive advantage, as does the present paper (Sirmon & Hitt, Reference Sirmon and Hitt2009; Sirmon et al., Reference Sirmon, Hitt, Ireland and Gilbert2011; Teece, Reference Teece2014a).

A competitive advantage is reflected in superior performance in capturing value and is indicated by ‘high relative profitability’ (Thomas, Reference Thomas1986: 3), ‘superior financial return’ (Ghemawat & Rivkin, Reference Ghemawat and Rivkin1999: 49), or ‘strictly positive differential profits in excess of opportunity costs’ (Foss & Knudsen, Reference Foss and Knudsen2003: 2). However, the RBV is not to ‘explain all types of profitability differentials’ (Peteraf & Barney, Reference Peteraf and Barney2003: 310). Instead, it is to explain only ‘long-lived differences in firm profitability’ attributable to heterogeneity in the way in which value is created using resources available, that is, heterogeneity in the productivity of all resources used in value creation (Peteraf & Barney, Reference Peteraf and Barney2003; Sirmon et al., Reference Sirmon, Hitt, Ireland and Gilbert2011; Teece, Reference Teece2014a). According to the RBV, superior productivity in value creation leads to long-lived superior financial gains which represent a sustained competitive advantage (Peteraf & Barney, Reference Peteraf and Barney2003; Teece, Reference Teece2014a).

Superior resource productivity can lead to long-lived superior financial gains for two reasons. First, superior productivity indicates that a firm can create more value with resources available and thus has more opportunities to capture a portion of the value created (Peteraf & Barney, Reference Peteraf and Barney2003). In a world of scarce resources, after all, it is the productivity of resources that ultimately determines the extent to which a firm captures the value it creates (Barney, Reference Barney1991, Reference Barney2001; Peteraf & Barney, Reference Peteraf and Barney2003). Second, superior productivity flows from superior capabilities, many of which are tacit and difficult to replicate (Sirmon & Hitt, Reference Sirmon and Hitt2009; Sirmon et al., Reference Sirmon, Hitt, Ireland and Gilbert2011). It is easy for rivals to imitate separate elements but hard for them to replicate the interlocked, coherent, and entire package. This is in line with the concept of causal ambiguity and social complexity (Barney, Reference Barney1991).

In essence, the RBV implies a framework in which capability building enhances productivity which, in turn, enhances financial performance. In other words, the capabilities undergirded by resource orchestration processes and practices generate an indirect effect on financial performance via productivity in addition to a direct effect on financial performance. A positive indirect effect indicates that capabilities contribute to long-lasting financial gains, while a positive direct effect indicates that these capabilities contribute to short-run financial gains contingent on ‘contextual factors’ (Peteraf & Barney, Reference Peteraf and Barney2003: 310). Firms should aim at long-lasting financial gains and develop capabilities to enhance the productivity of all resources to this end.

How does a firm enhance productivity? Generally speaking, firms can take two approaches to enhance productivity (Nishimizu & Page, Reference Nishimizu and Page1982; Färe, Grosskopf, Norris, & Zhang, Reference Färe, Grosskopf, Norris and Zhang1994; Coelli, Rao, O’Donnell, & Battese, Reference Coelli, Rao, O’Donnell and Battese2005). The two approaches can be elucidated in Figure 1 where all input resources are hypothetically divided into two bundles, and where the production frontier represents all possible resource combinations at which output is maximized at a given level of technology. The first aims to improve efficiency with which input resources are utilized using existent technology and know-how. This is indicated by the movement from points A, B, and C toward the production frontier F1 (Nishimizu & Page, Reference Nishimizu and Page1982). This portion of productivity increase was often called efficiency gains and can be achieved by imitating best management practices. The second approach aims at innovation to address new opportunities for resource utilization using novel technology and know-how and thereby push up the production frontier from F1 to F2 and then F3 (Färe et al., Reference Färe, Grosskopf, Norris and Zhang1994; Ghemawat & Rivkin, Reference Ghemawat and Rivkin1999). This portion of productivity increase was often referred to as technological progress in economics and cannot be achieved by simply imitating prevailing management practices (Coelli et al., Reference Coelli, Rao, O’Donnell and Battese2005). Accordingly, firms need to develop capabilities to orchestrate resources to enhance productivity in both ways. However, the specific capabilities firms need may differ depending on the context in which they operate (Mackey, Barney, & Dotson, Reference Mackey, Barney and Dotson2017).

Figure 1

Two sets of firm actions to enhance productivity. Note: All input resources are hypothetically divided into two bundles. The production frontier, F1, represents all possible resource combinations at which output is maximized at a given level of technology. One set of firm actions to enhance productivity is to improve efficiency in resource utilization using existent technology and know-how. This is often achieved by imitating best practices and indicated by the movement from points A, B, and C toward the production frontier F1 (solid arrows). The second set of firm actions to enhance productivity is to innovate to address new opportunities for resource utilization using novel technology and know-how and thereby push up the production frontier. This is indicated by the movement from F1 to F2 and then F3 (dotted arrows)

In advanced economies, most firms have reached a high level of operational efficiency and have to focus on the capabilities to engage in innovation to enhance productivity and, through it, profitability. In emerging economies, most firms face a notorious problem of operational inefficiency. In order to enhance productivity and, through it, profitability to attain a sustained competitive advantage in an increasingly competitive environment, they have to (1) implement best management practices to develop the capabilities to enhance operational efficiency in resource utilization using existent technology and know-how; (2) engage in innovation processes to develop the capabilities to embrace new opportunities for resource utilization using novel technology and know-how; and (3) develop the synthesis capabilities to do well at the same time. We therefore draw on the RBV to propose a model in which efficiency capabilities, innovation capabilities, and synthesis capabilities enhance productivity and, through it, financial gains. We illustrate our model in Figure 2. Differing from the efficiency framework of Bloom and colleagues, our model suggests that firms in emerging economies must develop the synthesis capabilities to combine efficiency and innovation to stay competitive (Bloom, Sadun, & Van Reenen, Reference Bloom, Sadun and Van Reenen2012; Bloom et al., Reference Bloom, Eifert, Mahajan, McKenzie and Roberts2013).

Figure 2

A theoretical model of firm capabilities and performance

Efficiency and firm performance

In advanced economies of North America, Western Europe, and Japan, for centuries, firms have been developing the capabilities to enhance efficiency in resource utilization in a highly competitive environment. Such capabilities are undergirded by efficiency management practices. Taylorism, Fordism, and Toyota lean production are among the well-known examples of such management practices (Bloom, Sadun, & Van Reenen, Reference Bloom, Sadun and Van Reenen2012). Despite criticisms, core elements of these best management practices have been accepted in most, if not all, businesses across advanced economies. There is very little room for firms to outperform rivals by exploiting the capabilities to further enhance operational efficiency. Bob Lutz (Reference Lutz2011), the former vice chairman of General Motors, made this point very clear to the automotive industry:

In order to keep rivals at bay, firms have to focus on innovation to address new opportunities for resource utilization using novel technology (Teece, Reference Teece2014a, Reference Teece2014b).

The picture looks quite different for firms based in emerging economies in Asia, Eastern Europe, Latin America, and Africa (Tian, Reference Tian1996; Buckley, Reference Buckley2009, Reference Buckley2011). Due to decades of isolation from competition, firms suffer from a lack of basic efficiency management practices. Bloom and colleagues conducted, for instance, efficiency management practice surveys across countries and found that emerging economies, such as Brazil, India, and China, had a large tail of very badly managed firms (Bloom, Sadun, & Van Reenen, Reference Bloom, Sadun and Van Reenen2012). In on-site visits to emerging economies, they found ‘firms without any formal maintenance program, inventory or quality control system, or factory organization’ (Bloom, Schweiger, & Van Reenen, Reference Bloom, Schweiger and Van Reenen2012: 594). To stay competitive, firms in emerging economies need to implement best management practices to build up the capabilities to overcome operational inefficiency.

As exemplified by the experience of firms in advanced economies, efficiency capabilities are undergirded by four sets of practices to manage resources. The first is target setting, which is related to the question of whether an organization supports long-term goals with tough but achievable short-term performance benchmarks. The second is monitoring, which is related to the question of whether an organization rigorously collects and analyzes operational performance data to identify areas in need of improvement. The third is problem-solving, which is related to the question of whether an organization promptly addresses problems in the value chain and makes sure that the problems will not happen again. The fourth is incentivizing, which is related to the question of whether an organization rewards high performers in operational efficiency with promotions and bonuses while retraining or removing underperformers. These practices are quite in line with what Bloom and colleagues called best management practices for operational efficiency (Bloom, Sadun, & Van Reenen, Reference Bloom, Sadun and Van Reenen2012; also Teece, Reference Teece2014a).

Efficiency improvement implies that firms are able to produce more output with a given amount of input resources and move further toward the production frontier. In firms based in emerging economies where efficiency management practices are lacking, the development of efficiency capabilities can foster productivity and, through it, financial gains and may constitute an important source of sustained competitive advantage (Nishimizu & Page, Reference Nishimizu and Page1982; Färe et al., Reference Färe, Grosskopf, Norris and Zhang1994; Peteraf & Barney, Reference Peteraf and Barney2003; Sirmon et al., Reference Sirmon, Hitt, Ireland and Gilbert2011; Teece, Reference Teece2014a). Indeed, Bloom and colleagues undertook a controlled experiment in India in which they provided free consultancy to 14 manufacturing plants on implementing these best management practices. After a year or so, the plants enhanced productivity by 17%, cut defects by more than 50%, reduced inventory by 20%, raised output by 10%, and increased profits to a different degree (Bloom, Sadun, & Van Reenen, Reference Bloom, Sadun and Van Reenen2012: 6; also Bloom et al., Reference Bloom, Eifert, Mahajan, McKenzie and Roberts2013). We thus propose the following hypotheses.

Innovation and firm performance

Generally speaking, innovation is related to the introduction of new ideas, methods, or things. Teece (Reference Teece2010a: 692–694) noted that the capabilities to innovate primarily consist of three key components: ‘1) identification and assessment of an opportunity (sensing), 2) mobilization of resources to address an opportunity and capture value from doing so (seizing), and 3) continued renewal (transforming). Innovation capabilities refer to the ability of a firm to engage in sensing, seizing, and transforming to address new opportunities’ (Chesbrough, Reference Chesbrough2010; Teece, Reference Teece2010b; Teece, Reference Teece2014a, Reference Teece2014b).

Firms based in different economies all need innovation capabilities but for different reasons. Firms in advanced economies have taken the lead in innovation for centuries and have developed a high level of technology and know-how. They are now in a position to rely on research and development (R&D) functions and professionals to focus on innovation in cutting-edge technology and breakthrough product designs, leaving most other parts of the value chain to firms in emerging economies through outsourcing, licensing, contract manufacturing, and other forms of strategic alliances (Buckley, Reference Buckley2009, Reference Buckley2011). As latecomers, in contrast, firms in emerging economies lag behind in innovation and have a low level of technology and know-how in almost all functional areas and need to engage in innovation in every part of the value chain to stay competitive in the global production networks (Williamson, Reference Williamson2010; Buckley, Reference Buckley2011; Williamson & Yin, Reference Williamson and Yin2014).

Therefore, firms in emerging economies have to develop innovation capabilities undergirded by innovation processes in all parts of the value chain. These innovation processes may involve (1) product innovation to offer products, which are often revised versions of breakthrough new products invented in advanced economies, to meet local customer needs; (2) process innovation to enhance product or service quality; (3) organizational innovation to enable a firm to respond to internal requirements and external pressures in an agile way;Footnote ² (4) marketing innovation to enable a firm to package its products, promote them, distribute them, and price them in ways that can address market needs or even create new markets; (5) logistics innovation to enable a firm to leverage ‘make-or-buy’ options as environments change; (6) externally contracted R&D in addition to in-house R&D in order to leverage knowledge sources unavailable within an organization; and, most importantly, (7) opportunities for all employees to develop and try out new ideas and approaches in any functional areas.

Buttressed by such processes, innovation capabilities enable a firm to engage in sensing, seizing, and transforming in different functional areas and generate novel technology and know-how to address new opportunities for resource utilization in the entire value chain. As shown in Figure 1, the novel technology and know-how allow a firm to push the production frontier upward. Innovation capabilities thus help improve productivity and, through it, financial performance (Nishimizu & Page, Reference Nishimizu and Page1982; Färe et al., Reference Färe, Grosskopf, Norris and Zhang1994; Peteraf & Barney, Reference Peteraf and Barney2003; Teece, Reference Teece2010a, Reference Teece2010b, Reference Teece2014a). We propose the following hypotheses.

Synthesis capability and firm performance

Synthesis capabilities refer to the capabilities to do two things simultaneously. A type of synthesis capabilities has been extensively discussed in the organization literature, that is, the ambidexterity to engage in exploitative and explorative innovation simultaneously (O’Reilly & Tushman, Reference O’Reilly and Tushman2004; Gupta, Smith, & Shalley, Reference Gupta, Smith and Shalley2006; Andriopoulos & Lewis, Reference Andriopoulos and Lewis2009; Gulati & Puranam, Reference Gulati and Puranam2009; Junni, Sarala, Taras, & Tarba, Reference Junni, Sarala, Taras and Tarba2013; Parida, Lahti, & Wincent, Reference Parida, Lahti and Wincent2016). Here we focus on the type of synthesis capabilities that firms in emerging economies need. We contend that faced with the dual pressures of efficiency and innovation, firms in emerging economies need to implement efficiency management practices while engaging in innovation and build the synthesis capabilities to do the two things equally well to enhance performance (Birkinshaw & Gupta, Reference Birkinshaw and Gupta2013; O’Reilly, & Tushman, Reference O’Reilly and Tushman2013; Fu, Flood, & Morris, Reference Fu, Flood and Morris2016). Synthesis capabilities of efficiency and innovation could be built up in a combined way or a specialized way depending on how efficiency capabilities and innovation capabilities are developed in a firm (He & Wong, Reference He and Wong2004; Cao, Gedajlovic, & Zhang, Reference Cao, Gedajlovic and Zhang2009). The two approaches to synthesis capabilities may vary in influencing productivity and, through it, financial performance.

A combined approach to synthesis capabilities involves a firm’s effort to increase the combined magnitude of both efficiency capabilities and innovation capabilities and focuses on their ‘absolute magnitude’ (Cao, Gedajlovic, & Zhang, Reference Cao, Gedajlovic and Zhang2009: 782). This approach implies that the two sets of capabilities are complementary to one another. The development of one set of capabilities can enhance the performance impact of the other. In contrast, a specialized approach to synthesis capabilities involves a firm’s effort to match the magnitude of efficiency capabilities with that of innovation capabilities and vice versa and focuses on ‘their relative magnitude’ (Cao, Gedajlovic, & Zhang, Reference Cao, Gedajlovic and Zhang2009: 782–783). This approach assumes that efficiency capabilities and innovation capabilities are ‘in opposition to each other’ and ‘orient the organization in the pursuit of different goals’ (Cao, Gedajlovic, & Zhang, Reference Cao, Gedajlovic and Zhang2009: 784; March, Reference March1991). The two sides must be closely matched to enhance firm performance. Clearly, which of the two approaches to synthesis capabilities help enhance firm performance is dependent on whether efficiency capabilities and innovation capabilities are complementary or conflicting.

We believe that efficiency capabilities and innovation capabilities are complementary, rather than conflicting (Cao, Gedajlovic, & Zhang, Reference Cao, Gedajlovic and Zhang2009). To build efficiency capabilities, for instance, a firm needs innovation capabilities to identify the opportunities and benefits which may flow from the development of such capabilities and creatively leverage these opportunities to reap the benefits. That is, they need to make sure that they do ‘the right things’ (Teece, Reference Teece2014a: 331). To turn the outcomes from innovation capabilities into productivity gains and financial benefits, similarly, a firm needs efficiency capabilities to set targets for capitalizing innovation outcomes in both the short-term and the long-term, to monitor the performance in innovation capitalization, and reward high performers and punish poor performers in the innovation capitalization process. That is, they need to ‘do things right’ (Teece, Reference Teece2014a: 331). Otherwise, innovation outcomes cannot translate into value to be created and captured by the firm due to operational inefficiency. Accordingly, synthesis capabilities are referred to as the combination of efficiency capabilities and innovation capabilities hereafter unless noted otherwise. If efficiency capabilities and innovation capabilities are complementary rather than contradictory, then a combined approach to synthesis capabilities is more likely to enhance productivity and, through it, financial performance than a specialized approach to synthesis capabilities (Birkinshaw & Gupta, Reference Birkinshaw and Gupta2013). We propose the following hypotheses.

Data source

We drew the sample from the raw data of the Enterprise Survey conducted by the World Bank together with the European Bank for Reconstruction and Development and the European Investment Bank for 36 emerging economies in Eastern Europe, Central Asia, Middle East, and Northern Africa in 2013–2015. These economies include Albania, Armenia, Azerbaijan, Belarus, Bosnia and Herzegovina, Bulgaria, Croatia, Czech Republic, Estonia, FYR Macedonia, Georgia, Hungary, Kazakhstan, Kosovo, Kyrgyz Republic, Latvia, Lithuania, Moldova, Mongolia, Montenegro, Poland, Romania, Russia, Serbia, Slovak Republic, Slovenia, Tajikistan, Turkey, Ukraine, Uzbekistan, Egypt, Jordan, Lebanon, Morocco, Tunisia, and Yemen.

The World Bank conducted the Enterprise Survey for other emerging economies as well over the period. The Enterprise Survey for the 36 countries was chosen because it included an innovation module which contained the information on innovation processes and management practices required to construct key variables in this paper. The Enterprise Survey for the 36 emerging economies started from Russia in 2012, followed by other economies over the 2012–2015 period. The Enterprise Survey was administrated for each country once only, so it is a cross-section data set. The data set contained 20,975 firms with no less than 20 employees. These firms were distributed in 14 industrial sectors, including food, wood, publishing and recorded media, chemicals, plastics and rubber, nonmetallic mineral products, fabricated metal products, machinery and equipment, electronics, precision instruments, furniture, other manufacturing, retail, and other services.

Measurement

Efficiency capabilities were proxied by four groups of efficiency management practices: target setting, performance monitoring, problem-solving, and incentivizing. As mentioned earlier, these management practices undergirded efficiency capabilities (Teece, Reference Teece2014a, Reference Teece2014b). The innovation module of the Enterprise Survey included a section on management practices which contained eight questions related to these management practices. As shown in Appendix 1, questions 1–3 were related to target setting; question 4 was related to performance monitoring; question 5 was related to problem-solving; and questions 6–8 were related to incentivizing. We first averaged the scores of questions 1–3 to construct a variable of target setting, and questions 6–8 to construct a variable of incentivizing. We then followed prior studies to average the scores of the four component variables to construct the variable of efficiency capabilities (Bloom, Sadun, & Van Reenen, Reference Bloom, Sadun and Van Reenen2012). A higher value represented a greater development of efficiency capabilities.

Innovation capabilities were undergirded by innovation processes in different functional areas and were measured on the basis of the questions in the innovation module of the Enterprise Survey, regarding whether the firm in the last 3 years (1) introduced new or significantly improved products or services; (2) introduced any new or significantly improved methods for production; (3) introduced any new or significantly improved organizational structures; (4) introduced new or significantly improved marketing methods; (5) spent on R&D activities, either in-house or contracted with other companies; (6) introduced any new or significantly improved logistical or business support processes; and (7) gave employees some time to develop or try out a new approach or new idea about products or services, business process, firm management, or marketing. We constructed a dichotomous variable for each answer to each of the seven questions, with 1 denoting yes and 0 denoting no. We then added the seven dichotomous variables together to construct a variable of innovation capabilities. A higher value represented a greater development of innovation capabilities.

Combined approach to synthesis capabilities was operationalized by multiplying efficiency capabilities and innovation capabilities (Cao, Gedajlovic, & Zhang, Reference Cao, Gedajlovic and Zhang2009). This measure reflects the combined magnitude of the two components (Gibson & Birkinshaw, Reference Gibson and Birkinshaw2004; He & Wong, Reference He and Wong2004; Cao, Gedajlovic, & Zhang, Reference Cao, Gedajlovic and Zhang2009). A higher value represented a higher level of the combined approach to synthesis capabilities. Specialized approach to synthesis capabilities was operationalized using the absolute difference between efficiency capabilities and innovation capabilities (He & Wong, Reference He and Wong2004; Cao, Gedajlovic, & Zhang, Reference Cao, Gedajlovic and Zhang2009).Footnote ³

Productivity was not directly observable. However, it could be measured using a production function proposed by Robert Solow (Reference Solow1956) – a Laureate of Nobel Prize in economics. The production function was illustrated in Equation 1:

(1) $$G_{i} {=}P_{i} S_{i} ^{{\beta 1}} A_{i} ^{{\beta 2}} $$

where _i represents firm. G represents the value of total sales revenues, S represents the number of total staff, and A represents the value of total assets. β1 and β ₂ represent marginal productivity of workforce and assets, respectively. Both are constants determined by available technology. P represents productivity (Solow, Reference Solow1956).

Taking the natural logarithm of Equation 1 produced Equation 2:

(2) $$LgG_{i} {=}a{+}\beta _{1} LgS_{i} {+}\beta _{2} LgA_{i} {+}\in_{i} $$

The constant a and the error term ∈ _i represent productivity (P _i), which was calculated using Equation 3:

(3) $$P_{i} {=}LgG_{i} {-}\beta _{1} LgS_{i} {-}\beta _{2} LgA_{i} $$

It is necessary to address the simultaneity bias and the selection bias in estimating labor coefficient (β ₁) and capital coefficient (β2) in Equation 2 (Yasar & Raciborski, Reference Yasar and Raciborski2008). Olley and Pakes (Reference Olley and Pakes1996) and Levinsohn and Petrin (Reference Levinsohn and Petrin2003), henceforth OP and LP, have developed two similar semi-parametric estimation procedures to overcome these biases using, respectively, investment and material costs as instruments for the unobservable productivity shocks. As data on investment were not available, we followed the LP procedure to use material costs as the instrument for the unobservable productivity shock in calculating productivity.

Financial performance was estimated using the accounting data provided in the Enterprise Survey. Specifically, profit was calculated as the difference between sales revenues and the costs of making a product or providing a service, including the cost of labor (salary, bonuses, and social security payments); the cost of raw materials and intermediate goods used in production; the cost of fuel and electricity; the cost of machinery, vehicles, and equipment; the cost of land and buildings; and other cost of production not included above. This is similar to the concept of gross profit in accounting. We divided profit by sales revenues to construct an estimate of returns on sales as a measure of financial performance. I also divided profit by assets to construct an estimate of returns on asset as an alternative measure of financial performance and used the variable in robustness test.

We took into account currency difference by transforming local currencies into US dollars. Following Bloom and colleagues, moreover, I transformed the dependent and independent variables into z scores by normalizing each variable to mean 0 and standard deviation 1 using the formula (Bloom, Schweiger, & Van Reenen, Reference Bloom, Schweiger and Van Reenen2012):

(4) $$z_{vi} {=}{{v_{{i{-}}} \bar{v}_{i} } \over {\sigma _{m} _{i} }}$$

where z _vi is the z score of the variable v _i in firm _i,$\bar{v}_{i} $ is the unweighted average of the variable v _i across all observations in all countries, and $\sigma _{m} _{i} $ is the standard deviation of the variable v _i across all observations in all countries. This transformation has two advantages. First, it helps minimize multicollinearity in using interaction terms in regressions. Second, it facilitates interpretation of the results as variables were measured in relative terms (Sirmon et al., Reference Sirmon, Hitt, Arregle and Campbell2010: 1387).

We included several control variables which have been considered to influence organizational ambidexterity and firm performance (He & Wong, Reference He and Wong2004; Andriopoulos & Lewis, Reference Andriopoulos and Lewis2009; Cao, Gedajlovic, & Zhang, Reference Cao, Gedajlovic and Zhang2009; Parida, Lahti, & Wincent, Reference Parida, Lahti and Wincent2016). We followed the Enterprise Survey to construct a dummy variable of firm size: 1 denoted large firms with more than 100 employees and 0 denoted small firms with less than 100 employees. We constructed a variable of firm age using the natural logarithm of the number of years since the firm was established. We constructed a variable of manager experience using the logarithm of the years in which the general manager had served on the position in any firms or companies. We constructed a dummy variable of foreign ownership: 1 denoted foreign firms with more than 20% of foreign share and 0 denoted domestic firms with less than 20% of foreign share. We constructed a variable of employee education using the logarithm of the average number of years of education of full-time employees. We constructed a dummy variable of employee training based on the survey question regarding whether the firm had formal training programs for its permanent, full-time employees in the last year. One denoted firms with an employee training program and zero denoted firms without such a program. We constructed a variable of product diversification using the percentage in the total sales revenues represented by the main product. We reversed the percentage so that a higher value indicated a higher level of product diversification. In addition, we constructed country dummies to control for variation in location, and industry dummies to control for variation in industrial affiliation. These control variables were included in all path analyses. The descriptive statistics and correlation of these variables are reported in Table 1.

Table 1

Mean, standard deviation, and correlation of variables^a

^a ** <.01; * <.05.

Before running path analysis, we need to address the problem of endogeneity. Theoretically, endogeneity is unlikely a problem for the variable of efficiency management practices because implementation of these practices, which did not require much financial resource, should be an antecedent rather than an outcome of an increase in productivity and profitability. It did present a problem for the variable of innovation processes since high-performing firms were likely to have more financial resource to engage in innovation. However, as the variable was calculated as an average over the previous 3 years, the problem of endogeneity was minimized. Indeed, Hausman test rejected the possibility of endogeneity for both efficiency management practices and innovation processes, with the coefficients of the estimated residuals from the reduced form regression being statistically insignificant from 0 at the .10 significant level.